Battlefield 6 - Anticheat Update - Season 3

Now that Season 3 is underway, and Ranked Mode has officially brought high-stakes competitive play to REDSEC, we wanted to provide a deeper dive into the ongoing efforts to maintain match integrity.  Competition increases attraction for bad actors. We have been busy analyzing the landscape, quietly infiltrating many of even the most private and exclusive cheat communities, building new detections and features throughout Season 2 that are now ready to be collectively unleashed.

Ranked Mode brings new high-stakes competitive play to REDSEC, and with that comes higher expectations from players for match integrity and fair play and an increase in attraction for cheaters to, well, cheat. Our teams have spent the second half of Season 2 finalizing new features, drafting up more stringent detections, and strategizing on when and where to deploy them. Some have already been active since the latter half of Season 2, others we’ve kept at the ready just waiting to turn the key. Let’s take a look at what new features we deployed specifically alongside the launch of Season 3, now that we don’t need the element of surprise against bad actors.

• New/Updated Cheating Detection Models – 

  • Behavioral Model Updates - We’ve created new machine learning models to better identify bad actors that are intentionally interfering with others’ gameplay beyond reasonable competitive behavior. These include actions like stream‑sniping, purposeful team sabotage, blocking spawns of vehicles or engaging in various forms of collusion or match manipulation, such as consistent informal team-ups, win-trading and other prohibited behaviors that go against fair play. We began seeing this disruptive conduct become more prevalent during the latter half of Season 2 and are now ready with trained models to help us identify and sanction such infractions.

  • Additional Botting Detections and Models - Starting in Season 1 and increasing in Season 2, we saw a rise in illicit botting services for Battlefield 6. Normal bot detections can often be easy to perform, but bad actors started to leverage cloud-gaming services to mask their use of adverse input devices and software that are the heart of these automated accounts. We’ve now deployed a cavalcade of new detections and investigative models, allowing us to effectively action a significant majority of the accounts and devices behind these illicit bots and their various infringing behaviours. From spamming advertisements for cheats/services, joining matches and intentionally sabotaging their team, joining various modes and feeding bad actors easy kills and victories in non-bot modes. We expect these detections to provide a significant reduction in the volume of unofficial bots going forward.

• New detections and Real-time enforcement acceleration -  Sometimes, when we have a really valuable detection, we don’t let it trigger the hammer immediately, these usually result in ban-waves of various sizes and cadences. While such techniques help us keep cheat developers from easily iterating and circumventing those detections with minimal effort, they also mean we are letting bad actors impact more players in more matches, and that isn’t ideal.  We now pair ban-wave queues with secondary rules and thresholds. When flagged accounts hit certain combinations of volume, reports, or game stats, enforcement triggers immediately instead of waiting for the scheduled wave. This has lowered our daily average time to enforcement.

• Security Feature Enforcement - 

  • TPM - We appreciate players who have already gone through the efforts of enabling secure boot. We were maintaining some allowances for players that were unable to enable it due to TPM chip issues, but we’ve been working with partners and various manufacturers to resolve these compliance issues.  TPM specifications are defined by the Trusted Computing Group here and formalized in the 2015 ISO requirements ISO/IEC 11889-1:2015. We are going to enable strict enforcement of the requirement going forward. 

    • For context, 98.76% of current players are already playing on hardware that is fully enabled and compliant with the title's requirements. Of the remaining 1.24%, the majority are actively using third-party emulators and/or spoofing their hardware compliance and/or are associated with some level of botting or cheat utilization efforts.

  • HVCI and VBS - These requirements have been present for the title since launch, but we haven’t enforced them yet. We are finalizing how we might deploy these to be most effective without negatively impacting legitimate players, and have not decided the most effective strategy yet. We may only require specific players with suspicious activity (like a lack of or falsified telemetry) and for specific modes like ranked. Or only for participation in official or sponsored tournaments, and are actively looking at the data and how cheaters behave to make these decisions.

As cheating communities adapt, so do we. This season, we’re excited to share that our anti‑cheat teams are expanding. With new members being added across our operations, engineering, and threat intel teams, giving us more resources and bandwidth to investigate cheating reports, create more sophisticated detections, develop anti-cheat features, gather accurate intel, and overall make sure we are unrelentingly protecting your gameplay experience.

As we’ve been monitoring various channels and social media regarding suspicions of cheating by players or claims regarding the false applications of enforcements, we are going to start to interact more directly with the community. We’ve created the BF_SledgeHammer account to allow our anti-cheat teams to engage in conversations on official channels more directly. Such subjects can be very sensitive, so we will only be engaging on the official EA Forums and Steam Discussions for Battlefield 6 as a pilot for this direct avenue of communication. 

Specifically, we are going to limit this interaction initially to posts claiming false positives or posting manipulated enforcement email contents. BF_SledgeHammer may decide to chime in to confirm such enforcement or provide some of the specific supporting details (like chat logs or the original account name) related to those enforcements that we believe will not undermine our detection efforts. We can’t promise to respond to all such posts or claims.

Understand that we cannot weigh in on the opposite, because any claim or support regarding the legitimacy of a specific account or player can immediately become false if that previously legitimate actor decides that today is the day they become a cheater, or such an account is compromised by a bad actor. 

To make it easier for players to identify bad actors, specifically confirmed cheaters, we are now going to additionally apply a VAC ban. We hope this will also help undermine the spread of misinformation regarding these bad actors.

Fair-play matters, and we aren’t going to relent, but that is all we are able to share right now regarding how we are working to maintain the fair and equitable experience for our players. We have lots of hammers still in the forge, hot and looking forward to sharing more details just as soon as they are tempered and ready to strike.

Stay sharp out there, soldier. See you on the Battlefield.